Echo State Networks as Novel Approach for Low-Cost Myoelectric Control

Published on Jun 21, 2017 in AIME (Artificial Intelligence in Medicine in Europe)
· DOI :10.1007/978-3-319-59758-4_40
Cosima Prahm5
Estimated H-index: 5
(Medical University of Vienna),
Alexander Schulz38
Estimated H-index: 38
(Bielefeld University)
+ 3 AuthorsGeorg Dorffner26
Estimated H-index: 26
(Medical University of Vienna)
Myoelectric signals (EMG) provide an intuitive and rapid interface for controlling technical devices, in particular bionic arm prostheses. However, inferring the intended movement from a surface EMG recording is a non-trivial pattern recognition task, especially if the data stems from low-cost sensors. At the same time, overly complex models are prohibited by strict speed, data parsimony and robustness requirements. As a compromise between high accuracy and strict requirements we propose to apply Echo State Networks (ESNs), which extend standard linear regression with (1) a memory and (2) nonlinearity. Results show that both features, memory and nonlinearity, independently as well as in conjunction, improve the prediction accuracy on simultaneous movements in two degrees of freedom (hand opening/closing and pronation/supination) recorded from four able-bodied participants using a low-cost 8-electrode-array. However, it was also shown that the model is still not sufficiently resistant to external disturbances such as electrode shift.
  • References (19)
  • Citations (2)
📖 Papers frequently viewed together
6 Authors (Cosima Prahm, ..., Georg Dorffner)
1 Author (David Hofmann)
4 Citations
4 Authors (Jiayuan He, ..., Ning Jiang)
2 Citations
78% of Scinapse members use related papers. After signing in, all features are FREE.
#1Khairul AnamH-Index: 7
#2Adel Al-Jumaily (UTS: University of Technology, Sydney)H-Index: 18
The success of myoelectric pattern recognition (M-PR) mostly relies on the features extracted and classifier employed. This paper proposes and evaluates a fast classifier, extreme learning machine (ELM), to classify individual and combined finger movements on amputees and non-amputees. ELM is a single hidden layer feed-forward network (SLFN) that avoids iterative learning by determining input weights randomly and output weights analytically. Therefore, it can accelerate the training time of SLFN...
10 CitationsSource
Jan 1, 2017 in ESANN (The European Symposium on Artificial Neural Networks)
#1Benjamin Paaßen (Citec)H-Index: 8
#2Alexander Schulz (Citec)H-Index: 38
Last. Barbara Hammer (Citec)H-Index: 34
view all 4 authors...
Modern bionic hand prostheses feature unprecedented functionality, permitting motion in multiple degrees of freedom (DoFs). However, conventional user interfaces allow for contolling only one DoF at a time. An intuitive, direct and simultaneous control of multiple DoFs requires machine learning models. Unfortunately, such models are not yet sufficiently robust to real-world disturbances, such as electrode shifts. We propose a novel expectation maximization approach for transfer learning to rapid...
3 Citations
#1Cosima Prahm (Medical University of Vienna)H-Index: 5
#2Benjamin Paaßen (Citec)H-Index: 8
Last. Oskar C. Aszmann (Medical University of Vienna)H-Index: 27
view all 5 authors...
For decades, researchers have attempted to provide patients with an intuitive method to control upper limb prostheses, enabling them to manipulate multiple degrees of freedom continuously and simultaneously using only simple myoelectric signals. However, such controlling schemes are still highly vulnerable to disturbances in the myoelectric signal, due to electrode shifts, posture changes, sweat, fatigue etc. Recent research has demonstrated that such robustness problems can be alleviated by rap...
12 CitationsSource
#1Ivan Vujaklija (GAU: University of Göttingen)H-Index: 13
#2Dario FarinaH-Index: 80
Last. Oskar C. Aszmann (Medical University of Vienna)H-Index: 27
view all 3 authors...
: Absence of an upper limb leads to severe impairments in everyday life, which can further influence the social and mental state. For these reasons, early developments in cosmetic and body-driven prostheses date some centuries ago, and they have been evolving ever since. Following the end of the Second World War, rapid developments in technology resulted in powered myoelectric hand prosthetics. In the years to come, these devices were common on the market, though they still suffered high user ab...
23 CitationsSource
#1Lizhi Pan (SJTU: Shanghai Jiao Tong University)H-Index: 5
#2Dingguo Zhang (SJTU: Shanghai Jiao Tong University)H-Index: 21
Last. Xiangyang Zhu (SJTU: Shanghai Jiao Tong University)H-Index: 26
view all 5 authors...
Background Most prosthetic myoelectric control studies have concentrated on low density (less than 16 electrodes, LD) electromyography (EMG) signals, due to its better clinical applicability and low computation complexity compared with high density (more than 16 electrodes, HD) EMG signals. Since HD EMG electrodes have been developed more conveniently to wear with respect to the previous versions recently, HD EMG signals become an alternative for myoelectric prostheses. The electrode shift, whic...
32 CitationsSource
Extreme learning machine (ELM) is a new learning algorithm for the single hidden layer feedforward neural networks. Compared with the conventional neural network learning algorithm it overcomes the...
Mar 23, 2015 in VR (IEEE Virtual Reality Conference)
#1Ivan Phelan (SHU: Sheffield Hallam University)H-Index: 2
#2Madelynne A. Arden (SHU: Sheffield Hallam University)H-Index: 19
Last. Chris Roast (SHU: Sheffield Hallam University)H-Index: 12
view all 4 authors...
Working together with health care professionals and a world leading bionic prosthetic maker we created a prototype that aims to decrease the time it takes for a transradial amputee to train how to use a Myoelectric prosthetic arm. Our research indicates that the Oculus Rift, Microsoft's Kinect and the Thalmic Labs Myo gesture control armband will allow us to create a unique, cost effective training tool that could be beneficial to amputee patients.
20 CitationsSource
#1Michael R. TuckerH-Index: 5
#2Jeremy Olivier (EPFL: École Polytechnique Fédérale de Lausanne)H-Index: 5
Last. Roger GassertH-Index: 33
view all 10 authors...
Technological advancements have led to the development of numerous wearable robotic devices for the physical assistance and restoration of human locomotion. While many challenges remain with respect to the mechanical design of such devices, it is at least equally challenging and important to develop strategies to control them in concert with the intentions of the user. This work reviews the state-of-the-art techniques for controlling portable active lower limb prosthetic and orthotic (P/O) devic...
318 CitationsSource
Dec 8, 2014 in NeurIPS (Neural Information Processing Systems)
#1Luca Pasa (UNIPD: University of Padua)H-Index: 3
#2Alessandro Sperduti (UNIPD: University of Padua)H-Index: 28
We propose a pre-training technique for recurrent neural networks based on linear autoencoder networks for sequences, i.e. linear dynamical systems modelling the target sequences. We start by giving a closed form solution for the definition of the optimal weights of a linear autoencoder given a training set of sequences. This solution, however, is computationally very demanding, so we suggest a procedure to get an approximate solution for a given number of hidden units. The weights obtained for ...
27 Citations
#1Dario Farina (GAU: University of Göttingen)H-Index: 80
#2Ning Jiang (GAU: University of Göttingen)H-Index: 31
Last. Oskar C. Aszmann (Medical University of Vienna)H-Index: 27
view all 7 authors...
Despite not recording directly from neural cells, the surface electromyogram (EMG) signal contains information on the neural drive to muscles, i.e, the spike trains of motor neurons. Using this property, myoelectric control consists of the recording of EMG signals for extracting control signals to command external devices, such as hand prostheses. In commercial control systems, the intensity of muscle activity is extracted from the EMG and used for single degrees of freedom activation (direct co...
290 CitationsSource
Cited By2
#1Cosima Prahm (Medical University of Vienna)H-Index: 5
#2Alexander SchulzH-Index: 38
Last. Oskar C. Aszmann (Medical University of Vienna)H-Index: 27
view all 8 authors...
Research on machine learning approaches for upper-limb prosthesis control has shown impressive progress. However, translating these results from the lab to patient’s everyday lives remains a challenge because advanced control schemes tend to break down under everyday disturbances, such as electrode shifts. Recently, it has been suggested to apply adaptive transfer learning to counteract electrode shifts using as little newly recorded training data as possible. In this paper, we present a novel, ...
Sep 28, 2016 in ICCCI (International Conference on Computational Collective Intelligence)
#1Martin TabakovH-Index: 2
#2Krzysztof FonalH-Index: 2
Last. Rami Qahwaji (University of Bradford)H-Index: 16
view all 4 authors...
In this paper a fuzzy model for control of bionic hand in real-time is proposed. The control process involves interpretation and analysis of surface electromyography signal (sEMG) acquired from patients with amputees. The work considers the use of force sensing resistor to achieve better control of the artificial hand. The classical type-1 Mamdani fuzzy control model is considered for this application. The conducted experiments show comparable results with respect to applied assumptions that giv...
4 CitationsSource